Contingency learning, i.e. learning that a cue predicts the presence (or absence) of an event, is central to the formation of beliefs regarding painfulness of body postures. Such beliefs may spread to safe cues due to compromised learning (e.g., excessive generalization, impaired safety learning), prompting avoidance and leading to disability. Despite its importance, compromised learning about low back pain is underinvestigated. We propose a low back pain scenario contingency learning task for the investigation of back pain-related learning.
Sixty healthy participants viewed pictures of an avatar in various back postures, and for each posture gave pain-expectancy judgments and viewed the verbal outcome (pain/no pain) for a fictive back pain patient. During acquisition, one posture was followed by pain (conditioned stimulus; CS+), whereas another was not (CS-). During generalization, unreinforced novel intermediate back postures (generalization stimuli; GSs) were tested. During extinction, only the CSs were presented, not followed by pain. During generalization of extinction, only the GSs were presented, not followed by pain.
Participants expected pain more for the CS + than the CS- (differential acquisition) and generalized their pain-expectancy to the GS most similar to the CS+ (generalization). During extinction, pain-expectancy for the CS + decreased and generalized to the GS most similar to the CS+ (generalization of extinction).
Future research should investigate generalizability of findings to clinical samples and consider the role of pre-existing pain threat beliefs.
This task is an easily applicable, non-invasive way to investigate the formation of back pain-related threat beliefs.

Copyright © 2020 The Authors. Published by Elsevier Ltd.. All rights reserved.